Engineering the Brain Immune System for Tumor Therapy Abstract Brain glioblastoma multiforme (GBM) kills through glioma cells that infiltrate normal surrounding brain, with consequent displacement and destruction of healthy brain tissue. Therefore, even aggressive surgical resection fails to excise completely all infiltrating cells and provide long term survival. Alternatively, the immune system could specifically target infiltrating tumor cells using specific immune mediated homing mechanisms. However, the immune system has reduced activity in the brain, due to the brain's immune privilege, i.e. the blood-brain barrier, a lack of proper antigen presenting cells (i.e., dendritic cells, DCs) in the healthy brain, absence of classical lymphatic outflow channels, and local expression of the endogenous immune inhibitors, i.e., transforming growth factor beta. Nevertheless, experimental and clinical studies have shown that immune-therapy can effectively eradicate experimental GBM, and can be harnessed to develop novel therapies for human brain tumors. Recently, we have shown that gene transfer mediated modification of the brain immune microenvironment induces systemic anti-tumor immune responses. Specifically, expression of fms-like tyrosine kinase 3 ligand (Flt3L) recruits antigen presenting cells (e.g., dendritic cells, DCs) to the brain parenchyma and into the tumor mass. To increase the availability of tumor antigens to the DCs, tumor cells are killed using conditional cytotoxicity [herpes simplex virus type 1 thymidine kinase (TK) plus ganciclovir (GCV)]. This combined approach induces tumor regression, long-lasting systemic immunological memory, and long-term survival in animals with large syngeneic tumors. The overall goals of the experiments outlined in this proposal are to define the immune cells involved in this anti-GBM immune response, determine the origin and functional activities of brain DCs, and elucidate the specific mechanism(s) underlying their ability to induce systemic long-term anti-tumor immunity. Our hypothesis is that Flt3L and conditional cytotoxicity recruit a specific subtype of dendritic cells to the brain, i.e. plasmacytoid dendritic cells (pDCs), which leads to a systemic anti-tumor immune response. Understanding the mechanisms that stimulate this immune response from within the brain in situ will lead to novel treatments for one of the deadliest human cancers. The novelty of our approach is our capacity to specifically recruit pDCs to the brain parenchyma and the tumor microenvironment in response to expression of Flt3L, an approach that achieves the priming of a systemic immune response against GBM from within the tumor mass in situ. We hypothesize that this will maximize the generation of an effective anti-GBM immune response to GBM antigens.